A 24/7 HVAC doctor-in-residence

Retragreen is an Australian start-up which has embedded AI into a product that enables rapid, remote energy monitoring at a granular level. The Retragreen unit combines hardware and software with a Bluetooth capability that gathers information from a BMS about HVAC energy use, lighting energy use and other systems, and then translates it into energy performance benchmarking and data.

According to the company, where a conventional energy audit involves a number of hours onsite by a consultant, this solution can be deployed for remote information gathering. The AI agent does the labour of crunching numbers, reducing the manual analytical workload, Lay explains.

“The AI agent is not a replacement for consultants, auditors, BMS teams or building managers. It is designed to do the heavy analytical lifting, support human decision-making, and make first-level HVAC insights more accessible and scalable,” he says.

“The AI is like a doctor for the asset owner, it can say ‘this is where the problem is,’ but it won’t do the treatment.”

HVAC is a specific focus, as it is one of the most significant energy loads in a building.

“This makes HVAC a useful entry point for energy reduction,” Lay explains.

HVAC data is also valuable for emissions reporting and for developing operational optimisation strategies and building retrofit and upgrade pathways.

Who is using it and how?

The technology has already been deployed for buildings in Malaysia, Thailand, Singapore and China. According to Retragreen, the data produced by the AI agent has helped building managers and consultants associated with those assets to achieve energy savings of up to 28%, depending on building type, operations and implementation approach.

These savings were achieved without capital outlay or equipment upgrades, Lay says.

The analysis delivered by the AI highlights where the operational wins might be, and where aspects of the HVAC system are inefficient or suboptimal.

The level of detail will also depend on the configuration of the building system, for example, whether a VSD in a chilled water system pump has a specific sensor that sends data to the BMS.

“Every piece of equipment in a building has its specifications for normal operational frequencies and energy use. The level of detail the asset owner or facilities manager inputs into the system will influence the level of detail the AI can access,” Lay says.

Real time, all the time

Beyond the use of the technology for energy efficiency and building analytics, Lay says there are also options for installing it as a 24/7 support for building management. This might look like the system sending alerts if building services performance deviates from expected set points, or if a system shows signs of breakdown or faults.

At the core of the technology solution is a real-time AI-enabled optimisation engine providing granular, system-wide visibility across HVAC operations which resolves performance down to component and zone level.

By continuously analysing variables such as equipment behavior, supply air conditions, and occupancy demand, the AI platform can also dynamically adjust airflow, temperature setpoints, and equipment sequencing to improve overall system efficiency.

---